<![CDATA[Teluk Bayur Port faces challenges of uncertainty and extreme fluctuations in general cargo volume. These fluctuations complicate the allocation of port facility resources and operational equipment. A comparison of forecasting methods was conducted to determine the best model for predicting cargo flow. The forecasting methods used were the Seasonal Autoregressive Integrated Moving Average (SARIMA) and Prophet. A quantitative descriptive approach was employed by analyzing monthly time series data from January 2023 to December 2025. The analysis results indicated that the data exhibited an upward trend and multiplicative seasonal characteristics. Model testing proved that the Prophet method with multiplicative seasonal adjustment provided superior accuracy compared to the SARIMA model ARIMA(2,1,2)(0,1,0)12. The MAPE error rate of the Prophet model reached 15.71% (classified as good), outperforming the SARIMA model at 24.48% (classified as reasonable). Furthermore, Prophet demonstrated better performance with an RMSE value of 1,682.66 tons and a MASE of 0.30. The Moving Range control chart test also confirmed the validity and stability of the prediction results for the Prophet method. The Prophet method proved to be more adaptive and stable in handling cargo flow fluctuations. This research fills the gap in port forecasting studies, which are generally dominated by container cargo, by applying algorithms to the dynamics of general cargo to create a tactical planning instrument for the management of Teluk Bayur Port.]]>
Copyrights © 2026
